Switch to optimum-intel internal testing models

tests/python_tests/data/models.py

@@ -6,7 +6,7 @@
 
 def get_models_list() -> tuple[str, ...]:
     model_ids: list[str] = [
-        "katuni4ka/tiny-random-phi3",
+        "optimum-intel-internal-testing/tiny-random-Phi3ForCausalLM",
     ]
     if pytest.selected_model_ids:
         model_ids = [model_id for model_id in model_ids if model_id in pytest.selected_model_ids.split(' ')]

tests/python_tests/data/tokenizer_configs.py

@@ -670,7 +670,7 @@ def get_tokenizer_configs():
             "unk_token": "<unk>",
             "chat_template": "{% for message in messages %}{% if message['role'] == 'system' %}{{'<|system|>'+ '\n' + message['content'] + '\n'}}{% elif message['role'] == 'user' %}{{'<|user|>' + '\n' + message['content'] + '\n'}}{% elif message['role'] == 'assistant' %}{{'<|assistant|>' + '\n' + message['content'] + '<|endoftext|>' + ('' if loop.last else '\n')}}{% endif %}{% endfor %}"
         },
-        "katuni4ka/tiny-random-phi3": {
+        "optimum-intel-internal-testing/tiny-random-Phi3ForCausalLM": {
             "bos_token": "<s>",
             "eos_token": "<|endoftext|>",
             "pad_token": "<|endoftext|>",
@@ -712,7 +712,7 @@ def get_tokenizer_configs():
             "unk_token": "<unk>",
             "chat_template": "{% for message in messages %}\n{% if message['role'] == 'user' %}\n{{ '### User:\n' + message['content'] }}\n{% elif message['role'] == 'system' %}\n{{ '### System:\n' + message['content'] }}\n{% elif message['role'] == 'assistant' %}\n{{ '### Assistant:\n'  + message['content'] }}\n{% endif %}\n{% if loop.last and add_generation_prompt %}\n{{ '### Assistant:' }}\n{% endif %}\n{% endfor %}"
         },
-        "katuni4ka/tiny-random-minicpm": {
+        "optimum-intel-internal-testing/tiny-random-minicpm": {
             "bos_token": "<s>",
             "eos_token": "</s>",
             "pad_token": None,

tests/python_tests/samples/conftest.py

@@ -94,7 +94,7 @@
         "convert_args": ['--trust-remote-code', '--weight-format', 'fp16']
     },
     "tiny-random-minicpmv-2_6": {
-        "name": "katuni4ka/tiny-random-minicpmv-2_6",
+        "name": "optimum-intel-internal-testing/tiny-random-minicpmv-2_6",
         "convert_args": ['--trust-remote-code', "--task", "image-text-to-text"]
     },
     "InternVL2-1B": {
@@ -118,15 +118,15 @@
         "convert_args": ["--task", "text-generation-with-past", "--weight-format", "int8"]
     },
     "tiny-random-latent-consistency": {
-        "name": "echarlaix/tiny-random-latent-consistency",
+        "name": "optimum-intel-internal-testing/tiny-random-latent-consistency",
         "convert_args": ['--trust-remote-code', '--weight-format', 'fp16']
     },
     "tiny-random-latent-consistency-lora": {
         "name": "katuni4ka/tiny-random-latent-consistency-lora",
         "convert_args": []
     },
     "tiny-random-llava": {
-        "name": "katuni4ka/tiny-random-llava",
+        "name": "optimum-intel-internal-testing/tiny-random-llava",
         "convert_args": ["--trust-remote-code", "--task", "image-text-to-text"]
     },
     "tiny-random-qwen2vl": {

tests/python_tests/test_llm_pipeline.py

@@ -202,14 +202,14 @@ def test_batch_string_inputs(
     )
 
 
-@pytest.mark.parametrize("llm_model", ['katuni4ka/tiny-random-phi3'], indirect=True)
+@pytest.mark.parametrize("llm_model", ["optimum-intel-internal-testing/tiny-random-Phi3ForCausalLM"], indirect=True)
 def test_batch_size_switch(ov_pipe: ov_genai.LLMPipeline) -> None:
     ov_pipe.generate(["a"], max_new_tokens=2)
     ov_pipe.generate(["1", "2"], max_new_tokens=2)
     ov_pipe.generate(["a"], max_new_tokens=2)
 
 
-@pytest.mark.parametrize("llm_model", ['katuni4ka/tiny-random-phi3'], indirect=True)
+@pytest.mark.parametrize("llm_model", ["optimum-intel-internal-testing/tiny-random-Phi3ForCausalLM"], indirect=True)
 def test_empty_encoded_inputs_throw(ov_pipe: ov_genai.LLMPipeline) -> None:
     with pytest.raises(RuntimeError):
         ov_pipe.generate(ov.Tensor(np.array([[]], dtype=np.int64)), max_new_tokens=2)
@@ -698,29 +698,35 @@ def test_pipeline_validates_generation_config(ov_pipe: ov_genai.LLMPipeline) ->
 # Work with Unicode in Python API
 #
 
-@pytest.mark.parametrize("llm_model", MODELS_LIST, indirect=True)
-def test_unicode_pybind_decoding_one_string(ov_pipe: ov_genai.LLMPipeline) -> None:
-    res_str = ov_pipe.generate(',', max_new_tokens=4, apply_chat_template=False)
+# Model, prompt and max_new_tokens that generate unfinished utf-8 string
+UNICODE_PYBIND_DECODING_TEST_CASES: list[tuple[str, str, int]] = [
+    ("optimum-intel-internal-testing/tiny-random-PhiForCausalLM", ",", 3)
+]
+
+
+@pytest.mark.parametrize("llm_model,prompt,max_new_tokens", UNICODE_PYBIND_DECODING_TEST_CASES, indirect=["llm_model"])
+def test_unicode_pybind_decoding_one_string(ov_pipe: ov_genai.LLMPipeline, prompt: str, max_new_tokens: int) -> None:
+    res_str = ov_pipe.generate(prompt, max_new_tokens=max_new_tokens, apply_chat_template=False)
     assert '�' == res_str[-1]
 
 
-@pytest.mark.parametrize("llm_model", MODELS_LIST, indirect=True)
-def test_unicode_pybind_decoding_batched(ov_pipe: ov_genai.LLMPipeline) -> None:
-    res_str = ov_pipe.generate([","], max_new_tokens=4, apply_chat_template=False)
+@pytest.mark.parametrize("llm_model,prompt,max_new_tokens", UNICODE_PYBIND_DECODING_TEST_CASES, indirect=["llm_model"])
+def test_unicode_pybind_decoding_batched(ov_pipe: ov_genai.LLMPipeline, prompt: str, max_new_tokens: int) -> None:
+    res_str = ov_pipe.generate([prompt], max_new_tokens=max_new_tokens, apply_chat_template=False)
     assert '�' == res_str.texts[0][-1]
 
 
-@pytest.mark.parametrize("llm_model", MODELS_LIST, indirect=True)
-def test_unicode_pybind_decoding_one_string_streamer(ov_pipe: ov_genai.LLMPipeline) -> None:
+@pytest.mark.parametrize("llm_model,prompt,max_new_tokens", UNICODE_PYBIND_DECODING_TEST_CASES, indirect=["llm_model"])
+def test_unicode_pybind_decoding_one_string_streamer(ov_pipe: ov_genai.LLMPipeline, prompt: str, max_new_tokens: int) -> None:
     res_str = []
-    ov_pipe.generate(",", max_new_tokens=4, apply_chat_template=False, streamer=lambda x: res_str.append(x))
+    ov_pipe.generate(prompt, max_new_tokens=max_new_tokens, apply_chat_template=False, streamer=lambda x: res_str.append(x))
     assert '�' == ''.join(res_str)[-1]
 
 #
 # Perf metrics
 #
 
-@pytest.mark.parametrize("llm_model", ['katuni4ka/tiny-random-gemma2'], indirect=True)
+@pytest.mark.parametrize("llm_model", ["optimum-intel-internal-testing/tiny-random-gemma2"], indirect=True)
 @pytest.mark.parametrize("generation_config,prompt", PERF_METRICS_TEST_CASES)
 @pytest.mark.parametrize("pipeline_type", [PipelineType.STATEFUL, PipelineType.PAGED_ATTENTION])
 def test_perf_metrics(
@@ -812,7 +818,7 @@ def test_perf_metrics(
     assert len(raw_metrics.m_durations) == num_generated_tokens - 1
 
 
-@pytest.mark.parametrize("llm_model", ['katuni4ka/tiny-random-gemma2'], indirect=True)
+@pytest.mark.parametrize("llm_model", ["optimum-intel-internal-testing/tiny-random-gemma2"], indirect=True)
 @pytest.mark.parametrize("generation_config,prompt", PERF_METRICS_STRUCTURED_OUTPUT_TEST_CASES)
 def test_perf_metrics_with_structured_output(
     ov_pipe: ov_genai.LLMPipeline,

tests/python_tests/test_parsers.py

@@ -25,7 +25,7 @@ def hf_ov_genai_models(request, tmp_path_factory):
 
 @pytest.mark.parametrize(
     "hf_ov_genai_models", 
-    ["katuni4ka/tiny-random-phi3"],  # this tokenizer is used as a stub only
+    ["optimum-intel-internal-testing/tiny-random-Phi3ForCausalLM"],  # this tokenizer is used as a stub only
     indirect=True
 )
 @pytest.mark.parametrize("answer", [
@@ -64,7 +64,7 @@ def write(self, message):
 
 @pytest.mark.parametrize(
     "hf_ov_genai_models", 
-    ["katuni4ka/tiny-random-phi3"],  # this tokenizer is used as a stub only
+    ["optimum-intel-internal-testing/tiny-random-Phi3ForCausalLM"],  # this tokenizer is used as a stub only
     indirect=True
 )
 def test_incremental_phi4_reason_integer_token_ids(hf_ov_genai_models):
@@ -92,7 +92,7 @@ def write(self, message):
 
 @pytest.mark.parametrize(
     "hf_ov_genai_models", 
-    ["katuni4ka/tiny-random-phi3"],  # this tokenizer is used as a stub only
+    ["optimum-intel-internal-testing/tiny-random-Phi3ForCausalLM"],  # this tokenizer is used as a stub only
     indirect=True
 )
 def test_incremental_integer_token_ids(hf_ov_genai_models):
@@ -144,7 +144,7 @@ def write(self, message):
 
 @pytest.mark.parametrize(
     "hf_ov_genai_models", 
-    ["katuni4ka/tiny-random-phi3"],
+    ["optimum-intel-internal-testing/tiny-random-Phi3ForCausalLM"],
     indirect=True
 )
 @pytest.mark.parametrize("split_answer", [
@@ -202,7 +202,7 @@ def test_incremental_phi4_reason_parser_nostreamer(answer):
 @pytest.mark.parametrize("do_reset", [False])
 @pytest.mark.parametrize(
     "hf_ov_genai_models", 
-    ["katuni4ka/tiny-random-phi3"],  # this tokenizer is used as a stub only
+    ["optimum-intel-internal-testing/tiny-random-Phi3ForCausalLM"],  # this tokenizer is used as a stub only
     indirect=True
 )
 @pytest.mark.parametrize("answer", [
@@ -266,7 +266,7 @@ def test_incremental_deepseek_parser():
 
 @pytest.mark.parametrize(
     "hf_ov_genai_models", 
-    ["katuni4ka/tiny-random-phi3"],
+    ["optimum-intel-internal-testing/tiny-random-Phi3ForCausalLM"],
     indirect=True
 )
 def test_custom_incremental_parser(hf_ov_genai_models):
@@ -308,7 +308,7 @@ def write(self, message):
 
 @pytest.mark.parametrize(
     "hf_ov_genai_models", 
-    ["katuni4ka/tiny-random-phi3"],
+    ["optimum-intel-internal-testing/tiny-random-Phi3ForCausalLM"],
     indirect=True
 )
 def test_final_parser_llama_32_json(hf_ov_genai_models):

tests/python_tests/test_sampling.py

@@ -21,8 +21,8 @@ def model_facebook_opt_125m() -> OVConvertedModelSchema:
 
 
 @pytest.fixture(scope="module")
-def model_katuni4ka_tiny_random_phi3() -> OVConvertedModelSchema:
-    model_id : str = "katuni4ka/tiny-random-phi3"
+def model_tiny_random_phi3() -> OVConvertedModelSchema:
+    model_id : str = "optimum-intel-internal-testing/tiny-random-Phi3ForCausalLM"
     return download_and_convert_model(model_id)
 
 
@@ -50,11 +50,11 @@ def model_tinyllama_1_1b_chat() -> OVConvertedModelSchema:
     ]
 )
 def test_basic_stop_criteria(
-    model_katuni4ka_tiny_random_phi3: OVConvertedModelSchema,
+    model_tiny_random_phi3: OVConvertedModelSchema,
     generation_config: GenerationConfig,
     prompt
 ):
-    generate_and_compare(model_katuni4ka_tiny_random_phi3, [prompt], generation_config)
+    generate_and_compare(model_tiny_random_phi3, [prompt], generation_config)
 
 
 @pytest.mark.parametrize(
@@ -159,14 +159,14 @@ def test_stop_strings_tinyllama(
     'I have an interview about product speccing with the company Weekend Health. Give me an example of a question they might ask with regards about a new feature'
 ])
 def test_greedy(
-    model_katuni4ka_tiny_random_phi3: OVConvertedModelSchema,
+    model_tiny_random_phi3: OVConvertedModelSchema,
     generation_config,
     prompt
 ):
     prompt = prompt.decode('unicode_escape') if isinstance(prompt, bytes) else prompt
 
     generate_and_compare(
-        model_katuni4ka_tiny_random_phi3, 
+        model_tiny_random_phi3, 
         prompt, 
         generation_config
     )

tests/python_tests/test_stateful_speculative_decoding.py

@@ -70,7 +70,7 @@ def test_string_inputs(main_model, main_device, draft_model, draft_device, promp
 def test_perf_metrics():
     import time
     start_time = time.perf_counter()
-    model_id = 'katuni4ka/tiny-random-gemma2'
+    model_id = "optimum-intel-internal-testing/tiny-random-gemma2"
     model_path = download_and_convert_model(model_id).models_path
 
     # Create OpenVINO GenAI pipeline:

tests/python_tests/test_structured_output.py

@@ -35,7 +35,7 @@ class RESTAPIResponse(BaseModel):
 
 structured_id_models = [
     "TinyLlama/TinyLlama-1.1B-Chat-v1.0",
-    "katuni4ka/tiny-random-phi3",
+    "optimum-intel-internal-testing/tiny-random-Phi3ForCausalLM",
 ]
 
 

tests/python_tests/test_text_streamer.py

@@ -88,7 +88,7 @@ def test_text_prompts(tmp_path, prompt, model_id):
     # '\n\n# 利用re.sub()方法，�' with UTF8 invalid for "microsoft/phi-1_5"
     [198, 198, 2, 10263, 230, 102, 18796, 101, 260, 13],
 
-    # '룅튜룅튜�' causes error on "openbmb/MiniCPM-o-2_6" / "katuni4ka/tiny-random-minicpmv-2_6"
+    # '룅튜룅튜�' causes error on "openbmb/MiniCPM-o-2_6" / "optimum-intel-internal-testing/tiny-random-minicpmv-2_6"
     [167, 96, 227, 169, 232, 250, 167, 96, 227, 169, 232, 250, 167]
 ]
 @pytest.mark.parametrize("model_id", tokenizer_model_ids)

tests/python_tests/test_tokenizer.py

@@ -296,7 +296,7 @@ def test_set_chat_template(ov_hf_tokenizers):
 @pytest.mark.parametrize(
     "ov_hf_tokenizers",
     [
-        "katuni4ka/tiny-random-phi3",
+        "optimum-intel-internal-testing/tiny-random-Phi3ForCausalLM",
         "TinyLlama/TinyLlama-1.1B-Chat-v1.0",
         # ("black-forest-labs/FLUX.1-dev", dict(subfolder="tokenizer")),  # FLUX.1-dev has tokenizer in subfolder
     ],
@@ -433,10 +433,10 @@ def hf_ov_genai_models(request, tmp_path_factory):
 @pytest.mark.parametrize(
     "hf_ov_genai_models",
     [
-        ("katuni4ka/tiny-random-phi3", {"padding_side": None}),
+        ("optimum-intel-internal-testing/tiny-random-Phi3ForCausalLM", {"padding_side": None}),
         ("TinyLlama/TinyLlama-1.1B-Chat-v1.0", {"padding_side": None}),
-        ("katuni4ka/tiny-random-llava-next", {"padding_side": "right"}),
-        ("katuni4ka/tiny-random-llava-next", {"padding_side": "left"}),
+        ("optimum-intel-internal-testing/tiny-random-llava-next", {"padding_side": "right"}),
+        ("optimum-intel-internal-testing/tiny-random-llava-next", {"padding_side": "left"}),
         (
             "BAAI/bge-small-en-v1.5",
             {"padding_side": None},
@@ -506,8 +506,8 @@ def test_padding(
 base_models_for_paired_input_test = [
     ("answerdotai/ModernBERT-base", {"padding_side": None}),
     ("TinyLlama/TinyLlama-1.1B-Chat-v1.0", {"padding_side": None}),
-    ("katuni4ka/tiny-random-llava-next", {"padding_side": "right"}),
-    ("katuni4ka/tiny-random-llava-next", {"padding_side": "left"}),
+    ("optimum-intel-internal-testing/tiny-random-llava-next", {"padding_side": "right"}),
+    ("optimum-intel-internal-testing/tiny-random-llava-next", {"padding_side": "left"}),
 ]
 
 def make_model_params():